Idea for doing multiple extrusion with one extruder

Warning: this is just an idea, it is untested at this time. I just had to describe it to you so I don't explode in a giant vapor cloud from my enthusiasm for what is possibly a very exciting technique of 3D printing. It goes like this, beginning with a picture that really helps explain it.

Attachment 3000

OK, on the left we have a support and on the right we have the desired part. Notice that the desired part is not printable with a single extrusion printer. Enter our new technique: we print the support with a support filament, then we let it cool while changing to our desired part filament, then we print our desired part. If the step up is not too high and a bit further apart than shown perhaps, we have accomplished dual extrusion with a single extrusion printer! The details of just how much of this sort of thing we can get away with are related to the geometry of your printer's extruder near the nozzle. All low-hanging parts must be considered such as the parts fan on my ROBO3D printer which may need to be removed to make this work.

I don't know what to call this technique, perhaps time division extrusion or stepped extrusion for an example. One could see usefulness of it in making printed circuit boards or in printing parts with gently curved undersides that would otherwise require support material that needs to be cut off the part, damaging the finish. And if you do have a dual extruder, fear not - it just became a quad extruder with this technique (or a hex or and octal for that matter). Please note that the part must have at least one location where it touches the bed (shown in glass in the image) because you can't start a print in mid-air (well you can with some finagling, but let's keep it simple for now).

Well anyway, just thought I'd pass along this idea in case anyone found it useful. Enjoy!

Les

G'day Les, I think this was covered about a year ago on the Reprap forums, I'll dig up the post.

Oh, thank you Geoff. With all the innovations bubbling up around the planet there is bound to be a good bit of duplication. I may use this technique to make circuit boards now that I think of it.

Les

Very interesting concept. Do you plan on trying it out Lester? I'd love to see the results!
Eddie

Yes Eddie, making music boxes. I hope to get some conductive ABS which is black in color to go with the white ordinary ABS I already have, then I will print test traces for starters. The first ones will be conductive ABS traces in various widths and thicknesses and then I will use this technique to embed those traces in a substrate. I know of a type of music electronics that a certain group of friends and I dabble in, pioneered by Stan Lunetta (a fumanchu-bearded Cali kook at heart) and carried on by us using new techniques on retro chips and new chips alike. I mention the CMOS music electronics because it is extremely low power such that traces of high resistance like 100 kOhms or more can actually be tolerated in most places. That's important because the conductive ABS is very high in resistance. Parts will be attached with conductive epoxy.

Quote:

---

Originally Posted by LesterCHall

Yes Eddie, making music boxes. I hope to get some conductive ABS which is black in color to go with the white ordinary ABS I already have, then I will print test traces for starters. The first ones will be conductive ABS traces in various widths and thicknesses and then I will use this technique to embed those traces in a substrate. I know of a type of music electronics that a certain group of friends and I dabble in, pioneered by Stan Lunetta (a fumanchu-bearded Cali kook at heart) and carried on by us using new techniques on retro chips and new chips alike. I mention the CMOS music electronics because it is extremely low power such that traces of high resistance like 100 kOhms or more can actually be tolerated in most places. That's important because the conductive ABS is very high in resistance. Parts will be attached with conductive epoxy.

---

If you find any conductive ABS that is actually capable of actual conductivity, please let me know! I have not found one that with a resistance of more than 10ohms.

These are some parts from my multi extrude extruder. It uses two nema 23 steppers and you can have as many bowden tubes as you want, this one has 7. Each one goes to a head and you don't change the filament, you just change the head. One stepper rotates until it is gripping the filament you want then the other rotates to extrude. Two steppers for as many colours as you want.

I haven't got round to making the head change automatic yet, it is all still a work in progress (on a back burner, once I proved it all to work I lost interest).

Attachment 3008Attachment 3009Attachment 3010Attachment 3011Attachment 3012

It is now in the public domain so Stratasys will probably patent it.

Well Geoff, the conductive ABS I've seen is specified at 1200 Ohm*cm. You multiply that by length and divide by cross-sectional area to get the resistance of a trace. I estimated that if really fat traces are used, then a 3 inch trace would be about 100 kOhms or lower. Most circuits cannot tolerate this much wiring resistance, but the very low power CMOS stuff can tolerate it. Also not all traces are 3 inches long, many are just 1/4 to 1/2 inch with careful design. If the material is reasonably consistent it should work for the intended application (music boxes).

Les

A simpler way to think of the resistance of a trace is to make your traces square in cross-section, then each cube is 1200 Ohms. So if your trace is 100 cubes long it will have a resistance of 120 kOhms, which is about what this type of circuitry can tolerate at max. It becomes an exercise of just counting cubes.

Les

Quote:

---

Originally Posted by LesterCHall

Well Geoff, the conductive ABS I've seen is specified at 1200 Ohm*cm. You multiply that by length and divide by cross-sectional area to get the resistance of a trace. I estimated that if really fat traces are used, then a 3 inch trace would be about 100 kOhms or lower. Most circuits cannot tolerate this much wiring resistance, but the very low power CMOS stuff can tolerate it. Also not all traces are 3 inches long, many are just 1/4 to 1/2 inch with careful design. If the material is reasonably consistent it should work for the intended application (music boxes).

Les

---

Good luck mate, sounds like you have it well thought out. I remember you from thingiverse when you made the Light2Heat converter back in May , it was impressive (Geoffro)

Ah, thanks for the encouragement Geoff, and glad you liked that design!

Les

there are a few new(ish) printers that use one nozzle and dual filaments.
Not sure how it works, but I've defintiely seen 2 companies with printers that work on that principle.
I'm presuming some kind of valve setup, no actual melting tube you'd probably need to melt within the nozzle itself - or use a lot of pruging between changing filament and clever retraction.

Unless you've got a delta printer it's probably a better option than printing one part and then printing another part seperately but inside it.

In cases like yours - it's probably easier to change the design to something easily printable without supports than to print twice.

Yes Aardvark, I am in agreement with you it sounds like a better option. The technique proposed here would only be useful in the event that your alternative was not a possibility.

Les